Abstract: Aiming at the problems of low quantization degree and strong social relevance in historical epidemic records, a historical epidemic classification method which based on simulated annealing and ant colony optimization is proposed in this paper. It uses a single ant to automatically generate the clustering result of epidemic data, and uses simulated annealing algorithm to optimize the clustering criteria, so as to obtain the global optimal solution of the epidemic clustering. In comparison with other clustering algorithms in performance, experimental results show that the proposed method has high accuracy and practicality.

Key words: historical epidemic classification, clustering, ant colony optimization, simulated annealing

摘要： 针对历史疫灾记录量化程度低、社会关联性强的问题，提出了一种结合模拟退火和蚂蚁算法的历史疫灾分级方法。利用单只蚂蚁对疫灾数据进行自动聚类并通过模拟退火算法对聚类准则进行优化，以获得疫灾聚类的全局最优解。通过与其他聚类方法的性能对比，实验结果证明该方法具有较高的精确性和实用性。

关键词: 历史疫灾分级, 聚类, 蚂蚁算法, 模拟退火算法